ISS research paper template



Graduate School of Development Studies

A Research Paper presented by:

Bahati Shabani Mlassah

(TANZANIA)

in partial fulfillment of the requirements for obtaining the degree of

MASTERS OF ARTS IN DEVELOPMENT STUDIES

Specialization:

[Economics of Development]

(ECD)

Members of the examining committee:

Prof. Arjun Bedi [Supervisor]

Prof. Michael Grimm [Reader]

The Hague, The Netherlands

November, 2010

Disclaimer:

This document represents part of the author’s study programme while at the Institute of Social Studies. The views stated therein are those of the author and not necessarily those of the Institute.

Inquiries:

Postal address: Institute of Social Studies

P.O. Box 29776

2502 LT The Hague

The Netherlands

Location: Kortenaerkade 12

2518 AX The Hague

The Netherlands

Telephone: +31 70 426 0460

Fax: +31 70 426 0799

DEDICATION

This work is dedicated to my Beloved Husband Amos L. Mgongolwa and my beloved children Faraja Amos Mgongolwa, Moses Amos Mgongolwa, and Beatrice Amos Mgongolwa whom with them wishing all the best for their life.

ACKNOWLEDGEMENT

It is not possible to mention individually all those who have provided some form of support in this work. However, it is my sincere belief that those concerned will accept these general acknowledgements.

Mentioning a few; I would like to thank Prof. Arjun Bedi my supervisor and Prof. Michael Grimm who was my second reader for their assistance from the very early stage in this research paper, and for the wise guidance they gave throughout this study. They always availed to me whenever I needed their help and made very constructive criticisms that helped to improve the quality of this work.

I wish to thank the staffs of National Bureau of Statistics (NBS) of Tanzania for their kindness by allowing for the access to the information that was important for this study to be successful.

Thanks for the Tanzania Government for provision of scholarship.

Finally but not the least; I am so grateful to my husband Amos L. Mgongolwa for his prayers, love, guidance, concern and tender encouragement , he considered as his concern to support and to boost my morale and overcome moments of hopelessness and unworthiness. In case of hardships he always made me regain my self-confidence, he was my motivator.

Contents

CHAPTER ONE: Introduction 1

1.1 Background 1

1.1.1 Role of the Public and Private Sector 3

1.1.2 Education system in Tanzania 6

1.1.3 Public expenditure on Education sector 7

1.2 Indication of the Research problem 8

1.3 Policy Relevance and Justification 9

1.4 Research Objective 10

1.5 Research Questions 11

1.6 Limitation of the study 11

CHAPTER TWO: Literature Review 13

2.1 Theory which guide the study 13

2.2 Empirical Literature Review 14

CHAPTER THREE: Discussion Methodology 20

3.1 The model 20

3.2 Data and Data Sources 22

3.3 Analytical framework 23

3.3.1 The schooling Model according to Mincer (1974) 23

3.3.2 Incorporating the Years in Schooling, Age and Sex in Earning Schooling Model 26

4.1 Data description for the year 1991/1992 28

4.2 Data description for the year 2000/2001 30

4.3 Data Description in 2006/2007 32

CHAPTER FIVE: Regression Results 37

5.1 Regression Results and Discussion for the year 1991/1992 37

5.2 Regression Results and Discussion for the year 2000/2001 42

5.3 Regression Results and discussion for the year 2006/2007 47

CHAPTER SIX: Analysis of Education Returns over time 51

6.1 Overall patterns of returns to education 51

6.2 Return to education by gender 51

CHAPTER SEVEN: Summary and Conclusion 56

References 59

Appendices 63

List of Tables

Table 1.1: Net Primary Education Enrollment from 1995 to 2009 2

Table 4.1: Education statistics for the year 1991/1992 28

Table 4.2:- statistical data description 2000/2001 30

Table 4.3 Data Description in 2006/2007 32

Table 4.4:- Education categories for 1991/1992, 2000/2001 and 2006/2007 presented in a summarized table for comparison purpose 34

Table 5.1: Regression results for the year 1991/1992 37

Table 5.2: log earnings functions estimates for Tanzania in 2000/2001 43

Table 5.3: log Earnings function estimates for Tanzania in 2006/2007 48

Table 6.1: Trend of Returns to education from 1991/1992 to 2006/2007 52

List of Figures

Figure 1.1: Distribution of Resources to Educational sub-sectors for the year 2008/2009 7

Figure 4.1: Education Statistics for the year 1991/1992 29

Figure 4.2: Summary Statistics for 2000/2001 30

Figure 4.3: Data Description for 2006/2007 33

Figure 4.4: Education categories for 1991/1992, 2000/2001 and 2006/2007 35

Figure 5.1: log earnings function estimates for Tanzania in 1991/1992 38

Figure 5.2: Return to education in 2000/2001 44

Figure 5.3 : Return to education in 2006/2007 in different education categories 49

Figure 6.1 Trend of Returns to education from 1991/1992 to 2006/2007 52

Figure 7.1: Trend of Return to education from 1991/1992 to 2006/2007 58

List of Acronyms

BEST Basic Education statistics in Tanzania

EFA Education For All

GDP Gross Domestic Product

HBS Household Budget Survey

HIV Human Immunodeficiency Virus

NER Net Enrollment Ratio

OLS Ordinary Least Square

PEDP Primary Education Development Plan

PSLE Primary School Leaving Examination

SEDP Secondary Education Development Plan

SAP Structural Adjustment Program

UPE Universal Primary Education

Abstract

Educational expansion in Tanzania following the education sector reform policy of 1995 has increased the enrollment rate at all levels of education primary, secondary to university and the number of trained young people in the labor market. Motivated by the education policy and the increase in educational expenditure, this study provides estimates of the private returns to different levels of education for the year 1991/1992, 2000/2001 and 2006/2007. The study uses data from the Tanzanian household budget survey for the mentioned years. Regressions of individual earnings on a set of explanatory variables including education and workings hours show that there are strong positive effects on individual earnings as the number of years of schooling increases. The highest returns are at the tertiary level while returns to primary education display a concave trend during the period. Also the study reveals that men had higher returns almost at all levels of education.

Hence higher return to tertiary education indicates opportunity for further investment. Therefore in order to benefit from the increased rate of return to tertiary education, government has to increase its investment on tertiary level as a means of allocating scarce resources in more productive investment which are highly paid in the labor market.

Relevance to Development Studies

Human capital is among the factors of production which has an influence on economic growth because it is through human capital investment in education that an individual acquires new knowledge on how to use new technology and therefore, investing in education has a positive effect on individual productivity and economic growth. Given the limited resources a country has, it is important to invest in education level which generates more return and this will be possible only through analyzing the return at different levels of education for decision making.

Keywords

Tanzania, Investment, Education, Return to education, Earnings, Human capital, Ordinary Least Square

CHAPTER ONE: Introduction

1.1 Background

Formally education in Tanzania was viewed as a public good and therefore only government who were responsible for the provision of education to its citizen. But due to financial constraints in 1980s Tanzania accepted the conditions of reducing its role from that of a key player to a facilitator this was through the structural adjustment program (SAP). As a continuation of SAP, in 1995 Tanzania introduced the education reform program which aimed to increase investment in education sector by involving both private and public sector. Following that reform policy both public and private sector have been investing more in education sector. For public sector about 18 to 19 per cent of the national budget is devoted to education sector which helped to increase school enrollment from level 55.9 per cent in 1995 to 95.96 per cent in 2009 for primary (see table 1.1) while for secondary education the net enrollment rate has increased from 5 per cent in 1995 to 23.5 per cent in 2008 (Ministry of Education and Vocational training, BEST 2009 pg 24 & 69). Despite the sharp increase in enrollment rates, Tanzania is still characterized by low levels of economic growth and standard of living unlike recently developed countries such as South Korea and Singapore where a part of their success is attributed to large investment in human capital (the East Asian Miracle: Economic growth and public policy 1993). Nevertheless education is still considered as a tool to raise individual earnings in developing countries. For policy makers it is important to investigate the returns to different levels of education and changes of educational returns over time as such information may be used to allocate resources in a more effective manner. The government of Tanzania realizes that quality education is the pillar of development because it is through education that the nation obtains manpower to serve in various sectors of the economy (http//tanzania.go.tz/education). Due to the important of skilled labor, it is believed that investment in education could create improved citizens and help to upgrade the general standard of living in a society. Also the positive relationship between education and earnings of a worker as highly educated and skilled workers tend to earn more than low skilled workers (Becker 1964) this encourage the individual to invest more for future earning hence the expansion of the national education system. Therefore government through education investment aim to improve national productivity and standard of living of its citizen while individual invest in education so as to generate future earnings

Table 1.1: Net Primary Education Enrollment from 1995 to 2009

|Primary Education |

|Primary schools Net Enrollment Ratio (NER) by sex and Gender parity index (GPI) |

|1995-2009 | | | | |

|Year |NER |

| |Male |Female |Total |GPI |

|1995 |55.9 |54.8 |55.4 |0.98 |

|1996 |55.9 |56.7 |56.3 |1.01 |

|1997 |58.1 |57.2 |56.7 |0.98 |

|1998 |56.0 |57.3 |56.7 |1.02 |

|1999 |56.4 |57.8 |57.1 |1.02 |

|2000 |58.6 |59.1 |58.8 |1.01 |

|2001 |65.8 |65.2 |65.5 |0.99 |

|2002 |82.1 |79.3 |80.7 |0.97 |

|2003 |90.4 |86.7 |88.5 |0.96 |

|2004 |91.4 |89.7 |90.5 |0.98 |

|2005 |95.6 |93.9 |94.8 |0.98 |

|2006 |96.8 |95.4 |96.1 |0.99 |

|2007 |97.6 |97 |97.3 |0.99 |

|2008 |97.5 |97 |97.2 |0.99 |

|2009 |95.85 |95.96 |95.90 |1.00 |

|Source:- Ministry of Education and Vocational Training (BEST 2009 pg 24) |

Also, Tanzania is among the countries which committed to the provision of education for all by 2015. Education For All (EFA) in Tanzania was a campaign focused on education financing in order to increase investment in education which helps to achieve the goal of education for all by 2015. EFA campaign was formally introduced in international conference held in Jomtien -Thailand in 1990 (Buchert 1997 pg 35). It was a move towards changing education focus from that of education for self- Reliance (socialist state) to the development of market economy which combines public and private initiatives. EFA was financed by World Bank followed the commitment done in Dakar by the World Bank President James D. Wolfensohn that Bank’s will help any country with viable and sustainable plan for achievement of education for All but without external finance to implement the policy (Mingat.et al 2002).

1 1.1.1 Role of the Public and Private Sector

The education sector reform aimed to change the government role from that of a key player to that of a facilitator in the provision of education. This new role of the government provides a more conducive environment for the private sector to increase its investment in education. From this, the nation believed that private investment in education will establish a more learning environment that will allow imparting both knowledge and technology to the youth for a more active participation in the agricultural sector and the economy as a whole (ibid). Following the education expansion, the enrollment ratio has increased from 54 per cent in 1990 to 95.9 per cent in 2009 for primary education for secondary was 5 per cent in 1995 to 26 per cent by 2008 (BEST 2009 pg 65). The increase in enrollment rate in secondary education is partly contributed by private sector (Non- Gov) as indicated in the table 1.2

|Year |Total primary |Enrolled to form 1 |Total |% |

| |sch. leavers | | | |

| |

Table 1.2: Form one Enrollment in public and private schools from 1995 to 2009

Enrollment rate for form one[1] peaked up in 2006 with total enrollment rate of 67.5% from 14.6% in 1995. The higher enrollment rate in 2006 was contributed by the implementation of secondary education development plan (SEDP1). SEDP is an education plan aimed to increase education expansion in underserved areas through reduction of school fees and construction of schools. The main aim of SEDP was to improve educational equity and reduce the poverty through human capital development at the secondary level for disadvantaged areas/groups. The effect of putting more efforts in secondary education () led to declining in quality of primary education as indicated in BEST 2006-2010. Here the declining in quality was assessed through the declining trend of pupils passing primary school leaving examination (PSLE). Furthermore the impact of decrease in number of pupils passing PSLE was observed by declining in secondary enrollment rate for three consecutive years from 2007 to 2009 as indicated in the table 2. Here the challenge may remain on how to increase form one enrollment rate without affecting the quality of primary education.

Having internalized the importance of education, Tanzania set the vision and mission of education which support the importance of investing more on education. Due to the importance of education, the vision of education in Tanzania states that “to have a nation with high level of education at all levels; a nation which produces the high quality and quantity of educated people sufficiently equipped with the requisite knowledge to solve the society’s problems, in order to meet the challenges of development and attain competitiveness at regional and global levels” (http//tanzania.go.tz). Following that vision, the ministry of education and culture has the mission to realize Universal Primary Education (UPE) aiming to eradicate illiteracy and attain a level of tertiary education and training commensurate with critical high quality human resources required to effectively respond to the development challenges at all levels (http//tanzania.go.tz/education). Also the importance of education for Tanzania’s development is recognized by incorporating in the National Strategy for Growth and Reduction of Poverty (NSGRP) which build three pillars; Growth and Reduction of Poverty, improvements of quality of life and social well-being; and Governance and Accountability.

2 1.1.2 Education system in Tanzania

Education goes far beyond, schooling. First “education is an essential human right, a force for social change — and the single most vital element in combating poverty, empowering women, safeguarding children from exploitative and hazardous labor and sexual exploitation, promoting human rights and democracy, protecting the environment and controlling population growth. Education is a path towards international peace and security” (Bellamy Unicef1999). Therefore education involves both formal and informal schooling although when we are dealing with education system we only consider formal education. From that argument, formal education system in Tanzania is divided into 5 parts, 2 years pre- primary, 1-7years is primary education (normally children starts primary school at the age of 7), followed by 4 years of secondary schooling leading to ordinary level and those who pass extend their schooling by 2 years leading to advanced level, and three years for bachelor’s degree, although some professions require more than 3 years of schooling like engineering. Secondary school and advanced level (post secondary) examinations results are used as criterion for selection of students for further formal education and training, and also for certification. Primary education in Tanzania is compulsory, and on average, children start standard one at the age of 7 which allow them to complete primary education at the age of 13 to 14 years.

Tanzanian education system has changed from that of colonial period because during the colonial era, the education system in Tanzania consisted of four years in primary school followed by four years of middle school, two of junior secondary and finally 2 years of senior secondary schooling. This education system started in mid 1930s to early 1960s few years after independent in 1961 (http//matokeo.necta.go.tz//history.htm, retrieved on 05 November 2010). Therefore, it is important to note that the former education system had an impact on categorization of education levels as those for example with 4 years of schooling were considered as incomplete primary education in our analysis while if we could be able to differentiate individuals with primary education from the older system it could lead to slightly differences in explanation concerning the education expansion in Tanzania

3 1.1.3 Public expenditure on Education sector

Generally the budget for education sector has been increasing over the last 8 years from 16.7 percent of total national budget in 2000/2001 to 19.8 percent in 2008/2009. The amount to invest in education sector depends on gross domestic product (GDP) of the country and the national priories. For example in 2000/2001 education sector received 2.7 as a percentage of GDP while in 2008/2009 education sector received 5.8 percent of the current GDP. Education sector is almost ranked as the first priority by giving large percentage of national budget compared to any other sector followed by heath sector (BEST 2010). Education sector is divided into four main categories the first category include primary, non-formal, other education institutions and supporting services. The second category includes secondary education while the third category comprises teacher education and the last category is tertiary and higher education. Following the mentioned four categories of education sector, higher resources are allocated to the first category primary non-formal, other education institutions and supporting services with 67.6 percentage share of the total budget allocated to education sector. The second category is tertiary and higher education which had a share of 21 percent followed by secondary education which account to 9 percent. Lower budget is allocated to teacher education which accounts to 2 percent as indicated in figure 1.1

| | |

|Figure 1.1: Distribution of Resources to Educational sub-sectors for the year 2008/2009 | | | |

|[pic] | | | |

| | | | |

| | | | |

|Source: BEST 2010 | | | |

| | | | |

|Government budget allocation to education sector is being enhanced annually in order to ensure better delivery | | | |

|of education services in terms of teaching and learning materials (ibid). However it is worth to know that | | | |

|education financing is shared among different stakeholders government, communities, parents and the student | | | |

|themselves | | | |

1.2 Indication of the Research problem

This study was influenced by Tanzania higher investment in education sector following the education reform policy of 1995 and economic growth rate of East Asian Tigers which experience high economic growth due to partly high investment in human capital unlike to Tanzania which allocate more resources to education sector but no improvement in economic growth. In 1995, Tanzania adopted the education sector reform which allowed more residents to get education including high level and tertiary education. Following that reform policy, more schools and universities established that has led to increase in enrollment in all levels of education within the economy. Since that time Tanzania allocates more resources in education sector with expectation that education will develop new knowledge and skills which will increase future productivity. And according to human capital theory, education is regarded as an investment like other investments although it can neither be liquidated nor reversible. This education policy is supported by Schultz (1988) by arguing that, if the relationships are causal, and education enhances the productivity and earnings of labor, it is not surprising that governments have been willing to expend a substantial fraction of national income on public education, neither it is hard to understand why parents have set aside an increasing amount of their private disposable income to educate their children, foregoing the productive contribution the children would have made to the family income[2] had they not attended school. The impact of increased in school enrollment on economic growth in Tanzania was analyzed by Seebens and Wobst (2005) by using a dynamic computable general equilibrium (DCGM) model, from their analysis they found that an increase in human capital formation in the long run leads only to a moderate increase of economic growth rates but a substantial improvement of factor incomes to low-education households at the same time as income effects are Pareto efficient. Then the effects of increase in enrollment on economic growth are of long run but the impact of schooling of individual earnings can be assessed after few years of schooling

1.3 Policy Relevance and Justification

Tanzania is among countries which found in south east of Sub Saharan Africa. Therefore the major characteristics which found in most of Sub Saharan Africa are also found in Tanzania. Tanzania is amongst the poorest countries in the world, it experiences low per capita income of only 500 US dollars in 2009 coupled with low standard of living to majority of its people (). The implementation of 1995 education reform policy for the intention of investing more in education by involving private sector, the outcome up to 2009 is that the economy of this country still characterized by low per capita income despite the enormous investment in education and the literature which support the policy. However, it is important to know that rapid economic growth of the eight high performing Asian economies (HPAEs) which include Hong Kong, the Republic of Korea, Singapore, Taiwan, China, Indonesia, Malaysia and Thailand was not only through accumulation of human capital but also physical capital and a more effective system of public administration facilitated the high level of domestic financial saving (International Bank for Reconstruction and Development 1993). Therefore due to increased budget allocation in education sector, it is important to have a study which will analyze the rationale of policy decision regarding the investment in education by investigating the returns on that investment. The study will focus on the impact of education on individual’s earnings.

Hence, this study is important because it provides information concerning education expansion in Tanzania and the effects of increased investment in education sector by analyzing the impacts of education expansion on changes in rate of return to different levels of education. Higher private return on a certain level of education will motivate households to invest more in that education level. For a developing country like Tanzania private return can be used to establish priorities for future education investment by allocating scarce resources on education level which yield more return. Therefore returns to education can be seen as indicator suggestive of area of concentration for investment (Psacharopoulos 2002). Although, it is not true that the education level which generate less returns are of less important to an individual or society or nation compared to the one which generate high returns.

1.4 Research Objective

The main subject of current significance of education expansion depends on the secular trends in rates of return to schooling. Students are completing their schooling with the assumptions of getting employment in the labor market while at the same time economists have normally assumed that the worldly increase in the relative supply of more educated persons will continue to reduce the gains from education. Therefore, the general objective of this paper is to analyze the effect of education expansion on earnings by estimating the returns to education in different levels of education by gender and also to assess the changes in returns over period of time from 1991/1992 to 2006/2007. Gender is a cross cutting issue and women are perceived to be weak in the labor market and they possess little capital due to past and present inequality of access to the capital market (Smith 1994). Since there is no study so far done in Tanzania concerning the return to education for men and women separately, therefore this study attempts to verify whether there is significant difference in returns between individuals with higher level of schooling and those with low level. Also to examine the difference in return between men and women. In order to be able to analyze this objective, the following specific objective were considered;-

- Analyze the trend of private returns to different levels of education by gender for the year 1991/1992, 2000/2001 and 2006/2007.

1

1.5 Research Questions

In order to meet the objective of the study, the research paper tried to answer the following questions

1. Do highly educated individuals receive higher rates of return to education compared to those with lower level of education?

2. Is there any difference in returns to education between women and men

3. Is there any difference in returns to education between rural and urban

4. What are the trends in education returns between 1990/1991 and 2006/2007, that is period before and after the 1995 education reform

1.6 Limitation of the study

The study only focused on the effect of education on individuals earnings. Also the returns to education are much more than individual earnings but due to measurement problem and difficulty of getting data the study focused only on monetary income earned by individual households. Absent and lack of important data concerning the cost of education the study concentrated more on private returns to education. The estimation of unity cost was not easily because different region has different enrollment ratio, which reflect that different region has different quality of education. For example, primary net enrollment ratio of pupils differs from one region to another, higher enrollment rate found in Kilimanjaro by 81% followed by Iringa 76% and Dar es Salaam 71% while low enrollment rate found in Lindi 44%, Shinyanga 46% and Kigoma 48%.

Also the estimation of the return to education may be overstated due to omission of some variables and the endogeneity of schooling. Although some approaches like IV can help to reduce the problem but this study only used OLS. It was easier said than done to calculate the return to non formal education. It was difficult to calculate the returns to non- formal education, since non- formal education refers to provision of education to a targeted group with a particular program outside the formal education system. This problem may lead to higher returns to a lower education level as the impact on income of non formal education was not recognized. Furthermore, it was possible that the estimation of female returns to education could be understated as non-marketed products were not taken into consideration on which large percentage of women time are spent on family work and non-market activities.

CHAPTER TWO: Literature Review

2.1 Theory which guide the study

The rate of return to education can be defined as a coefficient relating the percentage increase in earnings for the additional year of schooling (Card and Krueger (1996). According to investment theory as sited by Psacharopoulos 1973, “the rate of return on a project is a summary statistic describing the relation between the costs and benefits associated with the project”. From individual point of view, private return to education has a decision variable concerning the amount of schooling to receive. According to this study, return to education is determined by relating percentage change in individual earnings due to increase in one year of schooling.

From the definition of rate of return to education; the theory which guides our analysis is human capital theory. The essence of using human capital approach rests on the assumption that individuals try to make decision on how to optimize judgments regarding the acquisition of further personal characteristics that yield net benefit in the future times (Fallon and Verry 1988). Human capital investments include anything that generates an individual self improvement which includes health, education and post-school investment in training, although more attention has been directed towards education (ibid). Therefore Human capital theory linked the education investment in its achievement with individual productivity and earnings. The suggested relationship was put into a summarized form as indicated below:-

|Education Human capital Productivity Earnings. |

Knowledge embedded in human capital play an important role in labor productivity and hence on growth path of general economic activities which finally command for higher earnings in the labor market. For example, consider the production function[pic], where output (Q) depends on factor labor (L), capital (K) and raw material (R). Output can only increase by increasing the amount of factor inputs or by increasing the productivity of those inputs (Smith 1994). Also Mincer (1974) emphasized the positive relationship between individual’s schooling and his/her subsequent earnings as individuals’ level of education influences the productivity capacity and earnings.

For that case the optimal decision regarding further investment on human capital depends on the decision regarding the individual maximization of total wealth where total wealth is maximized only by involving human and non human capital components. Hence the individual has to compare the return to investment on different investment alternatives (Fallon and Verry 1988). Also the return to schooling can be compared with the market interest rate to determine the optimal investment in human capital (Heckman et al. 2003, Schultz 1988, Psacharopoulos 1973)

First education is considered as an engine of the economic growth, it is through education a nation can meet challenges of the economic growth by increasing knowledge and productivity which enable to compete with other countries in the world market. Griliches (1988) used Cobb-Douglas production function to test the significance of education variable in agricultural studies and found that there were variations in output per farm although the output did not vary greatly. Hence the role of education in providing knowledge, skills and technology to enhance production cannot be ignored.

2.2 Empirical Literature Review

There are number of studies concerning factors which determine the level of individual earnings, but in order to fit our study it worth to describe some of them as those described will provide evidences of our study findings. Bedi and Edwards (2002) did a study on the impact of school quality on earnings and educational returns in low income countries. The result from his study showed that, ‘men educated in counties with better quality schooling earned significantly higher incomes than those men educated in low- quality schools’. This means that people of the same level of education will have different return to education provided they are from different school quality background.

Schultz (2004) used the sample of six developing countries from which he found that, private returns in six countries are highest at the secondary and post secondary levels which was higher for women compared to men. He also observed that the wage structure by schooling, age and sex does not represent an accurate picture of the productive gains that would be realized by increasing the supply of better educated individuals. He complained the issue of high annual increment on wages which may not reflect the productive capacity of an individual. Also the findings of high return on higher levels of education was observed by Aromolaran (2004), who did study on wage returns to schooling in Nigeria. His results show that the return to primary education is 2.5 percent for men and 2.4 percent for women, secondary level has a return of 3.9 percent for men and 4.4 percent for women and post secondary is 10.4 percent for men and 12.2 percent for women. Therefore, this result shows a positive relationship between increase in level of education and the returns to education. But the results by Psacharopoulos (1994), finds that the return to schooling are higher in primary level, while it decreases as the number of year in schooling increases. Also Psacharopoulos results show that the private return is higher than the social returns.

The study on the returns to investment in education which was conducted by Psacharopoulos (2002), came with the argument that returns to schooling tends to decline as average schooling level increases. His findings correlate with that of 1994, from that debate it is difficult to generalize that returns to education are higher in low level and tend to decline as average schooling increases due to contradicting results by different studies. The result are different due to may be different market situation basing on demand and supply theory increase in supply of labor compared to demand will lower the price that labor which result to decline in return to education regardless the education level. The important of market forces to determine price is noted by Mincer (2003) who argued that ‘increasing supply of human capital cumulates overtime in the form of training and school enrollments, eventually reducing the wage gap equilibriums, that is to the rates of return comparable to those on alternative investments’. Decrease in wage gap will lead to more employment of highly educated workers and create unemployment to low educated individuals. Becker (1964) and Smith (1994, 2003) Support the statement of more employment of highly educated workers by arguing that unemployment tends to be strongly correlated but usually inversely related to education levels. Therefore inequality in distribution of earnings is usually associated with inequality of education levels and training

Also age is another factor which can determine earnings indirectly. Because there are some factors behind age which affect the level of individual earnings like the individual level of productivity. Becker 1964 continued by arguing that earnings tend increase with age at a decreasing rate, whether the rate of increase or retardation all are positively related to individuals level of skills (ibid). What makes individual with younger age to earn more income is the level of individual flexibility, younger persons has a chance of receiving more schooling and training depending on the market demand compared to older ones and therefore increasing a chance of getting highly paid employment or jobs. In addition factors which contribute to the declining of income of older workers including the obsolescence of education and skills and sometimes decision to work few hours (Thornton et al 1997). In addition sometimes data itself may be a source of showing declining in income in older ages because the data do not track the earnings of specific individual through their life span (ibid). But the problem with age-earning profile is that it works only under assumption that there is no period effect on earnings like inflation and the person’s productivity remain constant for the whole lifetime which actually may not hold. Therefore age earnings profile could be worth enough if it could trace the earnings of specific individual through their lifetimes.

Wilmoth and Koso (2002) did a study on the impact of marital status on wealth. Their result showed that individuals who are not continuously married have significantly lower wealth than those who remain married throughout the life course. Remarriage offsets the negative effect of marital dissolution. There are significant gender differences in these effects. Men experience short term increase in resources while for women economic resources drop approximately one -third during the first year after divorce and the economic loss persists until remarriage. Therefore the sequence of marital events provides a detailed picture of the life paths that lead to wealth heterogeneity among the older population. Married older couples had higher median incomes and net worth than older adults who were widowed, divorced or never married (Seigel, 1993) as quoted by Wilmoth and Koso (2002).

In addition, Psacharopoulus (1973) in the study of returns to education, he came with conclusions that increase in the proportional of higher education graduates relative to secondary school graduates in the labor force will lead to fall in the relative wages, and this will lead to more employments of more university graduate relative to secondary school graduates in the economy. Psacharopoulus argument was linked with competitive market, but under monopolists it will depend on the market conditions and the type of labor because sometimes it is difficult to substitute labor with higher level of education due to difference in professions and normally employers have less incentive to fire employees with specific qualification which fits the firm compared with those without qualification

Segundo (2003), study on family background and returns to schooling in Spain, the results shows that; socio- economic background of the family affects not only the accumulation of human capital on the part of individuals, but also the returns that this investment produces in the labor market. The results shows that children of illiterate parents their returns are around 3.3 per cent per year of schooling, while for the children of parents with primary schooling the figure is 5.7 per cent reaching 7.3 per cent for those who parents completed secondary education and higher. Segundo (2003) did not look the source of bias in intergroup comparison like different in participation in the labor force which has a great on ones earnings.

Despite many studies concerning factors which influence individual earnings, Diaz (2008) did a study on comparison between urban and rural return to education. Diaz found that investing in education is profitable in both areas, and the returns to education were found to be higher in rural area compared to urban in every level of education. The return to education for women was higher in basic education levels while for men at higher levels of education. Education in urban proved more profitable for men at primary and higher education levels while for women in some years at lower level and upper of education.

The study done by Kahyarara (2006) from Tanzania has added new evidence to this debate by comparing returns to vocational and general education of workers in Tanzanian manufacturing firms. The evidences were based on cross sectional data in which provided a comparison of the returns to general and vocational education using firm level panel data with substantial information that allowed a control for time invariant firm attributes, endogeneity of education and other worker- firm characteristics. The findings were; general education was more rewarding than vocational education and on the job training. The marginal rate of returns to one year of education was ranging between 4.8 and 17.5 percent compared to the rates of returns to one year of vocational education that ranges between 1.4 and 2.8 percent. The results were stable even after controlling for endogeneity, firm-worker characteristics and firm fixed effects. The finding showed that at high level of general education unemployment in the Tanzanian labor market was low.

Also the study done by Seebens and Wobst (2005) on the impact of increased school enrolment on economic growth in Tanzania asserts that the wage gap between unskilled and skilled labor categories was narrowing, the closing of the wage gap mainly applies to low wage for non-educated labor and medium wages paid to primary educated labor, where the latter are still much lower than the wages paid to secondary educated labor. The study also revealed that a female primary educated earned 3.5 times as much as the non educated female worker, which was a large decline as compared to 5.2 times in the base year. On the other side; the male primary non-educated, the study revealed that the labor earnings ratio dropped from 2.8 to 1.3, this signified that the wage gap for these categories dropped by more than a half, from that the paper concluded as increase in production of primary educated labor led to a decrease of its marginal productivity followed by increase in supply of labor force with primary education.

In addition, the study conducted on The Dynamics of Returns to Education in Kenyan and Tanzanian Manufacturing came with the findings that Kenya had seen long-run fall in the return to education at the post primary level and Tanzania had an increase in the return to education in the 1990s (Söderbom et al. 2005). The study also found that n Tanzania average marginal returns rose from 6 per cent in 1993 to 9 per cent in 2001 for the young, and from 8 to 13 per cent for the old, for both countries, average returns were always higher in the old age group, for both countries and both age groups the marginal return on post-secondary education was higher, in most cases substantially higher than the returns before the secondary level. In both countries and for both age groups the average marginal return on education for those with postsecondary exceeded 27 per cent in 2000/01. It was evidenced that a large and statistically significant increase in the return to post-secondary education in Tanzania between 1993 and 2001. Thus, the main reason for such increase in the average return in Tanzania documented above was due to the earnings function that has become significantly more convex. On total, wage differentials were associated with various levels of education over a fixed baseline level. In Kenya, it was evidenced that the differences in the wage earnings across age groups was due to relatively high levels of education. On balance, earnings differentials attributable to education are larger in the old age group than in the young.

CHAPTER THREE: Discussion Methodology

3.1 The model

Education is considered as an investment like other investments in which existing resources are invested for future earnings. Education plays a great and significant role to an individual and to the economy of a nation as whole, consequently expenditures on education are found to be a form of investment for future earnings. This leads to expansion of an individual’s human capital and increases the chance of employment and thus enhances earnings for the individual worker. Therefore the optimal investment decision is found where an investment in the [pic]year of schooling offers an internal rate of return higher than the market rate of interest (Heckman et al 2003, Fallon & Verry 1988). If the individual earning is constant through the working life and the cost of education is zero then the Mincerian model of 1974 can be used to estimate the private return to education for both female and males. The empirical specification treats log earnings as a function of different levels of education and includes the other regressors that may have a bearing on earnings. By following Mincer (1974) model, we were able to estimate the marginal return to education by using the logarithmic equation as indicated below:

[pic]

Where:- [pic] is the monthly earnings for an individual[pic], [pic] is a constant which represents the average earning for an individual who has zero level of education. The coefficient [pic] measures the extent to which schooling level rises earnings above the reference level of education; while [pic] is a dummy variable representing the highest level of education attained by an individual; [pic] is age while [pic] is a negative coefficient of the quadratic experience(age) term which produces an age earnings profile that is concave from below. Therefore ‘age can be viewed as an inherent depreciation phenomenon in the human capital terminology’ (Mincer 1974). And [pic] is a vector of control variable such as sex and marital status. The “[pic]’ error term is added to capture the role of unobserved characteristics in determining individual earnings and is assumed to be uncorrelated with all variable in the right hand. According to Mincer (1974) experience is among the determinants of log earning function under the notion of learning by doing, but according to unavoidable situation the data does not show the experience profiles therefore the return to education may be overstated due to omission of such important variable of experience although the inclusion of age and age squared may help to reduce the problem. Also the information concerning number of days worked per month and earning per hour for each individual was available and is included in the specification. Also in order to capture regional labor market differences, two dummy variables for urban and rural areas were included, although the country has twenty regions but the aim of the study is to estimate the return to different levels of education by considering gender and location targeting to rural or urban.

The Mincerian model is applied under the following assumptions; firstly individual earning is constant throughout the working life. The second one is there are zero depreciation during the schooling period and the last assumption is there is no or zero investment during the working period.

From those assumption the Private marginal return was calculated by using the formula [pic] where [pic] is equal the difference in regression coefficient of two education categories, while [pic] stands for difference number of years of schooling between the two level of education. The difference in coefficients divided by the number of years of schooling separating the two groups the result was multiplied by 100 to convert into percentage. Also the years of schooling for incomplete levels of education were obtained by calculating the average years of schooling, for example for incomplete secondary education there were individuals with eight years of schooling, nine and ten, where by the average taken was nine= [pic] the same procedure was applied for incomplete primary education.

The estimates of rates of return associated with different levels of education was estimated under the assumption that individual with higher level of education will be more productive that one with low level of education and therefore the private rate of return will be positively correlated with the level of education. The main reason of using the log of earnings was to make suitable transformations of variables which are nonlinear but are linear in parameters in which the transformed variables can be easily estimated by using OLS regression. The second reason is to impose a constant percentage effect of education on earnings (Gujarati 2009). Also the log earning allows the investment variables to be expressed in units of time like years of schooling and years of experience Mincer (1974).

3.2 Data and Data Sources

The secondary data were collected from different sources including Tanzania Bureau of Statistics (NBS), Ministry of education and vocation training. The data from National Bureau of statistics (NBS) concerning individual earnings, working hours and education levels on each individual’s level of education by age and sex was collected by NBS from twenty regions in Tanzania mainland (Zanzibar was excluded). The data for 1991/92, 2000/2001 and 2006/2007 were collected which helped to analyze the trend of private return to different levels of education. The data from 1995 to 2009 concerning student’s enrollment rate in education from primary level to secondary level were purposely collected from ministry of education and vocational training as those information helped to access the expansion of education system in Tanzania following the education reform program of 1995.

The data files used in analysis in all three data sets were created by considering only individual who are working in rural or urban and whether he/she was employed in formal or informal sector. The sample included male and female so as to substantiate the difference in private rate of return between males and females. In order to obtain the total earnings of each individual, different sources of earnings were summed for each individual separately. The total number of hours worked was obtained by summing up the number of hours spent in main economic activities and on secondary activities.

Different levels of education were identified by considering the number of years in schooling. Individual with zero years of schooling were termed as those without education, one to six years of schooling as incomplete primary education according to Tanzania education system while with 7 years of schooling were considered as complete primary education. Eight to ten years of schooling were considered as incomplete secondary education and those with eleven years of schooling were considered as complete secondary education. Advanced level was awarded to those with thirteen years of schooling and diploma those with 15 years of schooling. Tertiary education is the highest level of education an individual attained if he/she had 16 and above years of schooling.

A number of dummies were constructed including that of education levels, urban or rural and sexes, we put 1 for male and 0 for female. Urban 1 and rural 0

3.3 Analytical framework

1 3.3.1 The schooling Model according to Mincer (1974)

In calculating the effects of schooling on the earnings, it is first assumed that postponement of the earnings due to the lengthier schooling is tantamount to a reduction of the earning span.

Let: n = the length of the working life plus length of schooling= the length of the working life for persons without schooling.

Ys = Annual earning of an individual with s years of schooling.

Vs = Present value of an individual’s life time earnings at start of the schooling.

R = Discount rate the individual uses to discount the future

t = 0, 1, 2,…….., n time in years.

d = difference in the amount of schooling in years

e = base of natural logarithm

Then a discounting expression can be written as:

[pic]……………………………………………………1

The process of discounting is discrete, for the process to be continuous the expression could be written as

[pic]……………………………………..2

Again considering the present value of life time earnings of an individual of who engaged in s-d years of schooling is:

[pic]…………………………………………3

The ratio, [pic], of annual earnings after s years to earnings after s-d years of schooling is found by letting [pic]……………..…….………………………………………..4

[pic]………………………..…5

Looking at the equation it can be easily seen that ks,s-d possesses the following characteristics;

Larger than a unity, a positive function discount rate r and negative function of n

Following the above observation people with more schooling have higher annual pay, the difference between earnings of individuals due to the difference in the investment of d years of schooling is larger the higher the rate of return on schooling, also the difference is larger the shorter the general span of working life, since the costs of schooling must be recouped over a relatively shorter period.

By considering the effects of relationship between s and n on the value of ks,s-d for when s (d fixed), that is the relative income differences between, say 10 and 8 years of schooling are larger than those between individuals with 4 and 2years of schooling respectively. But when n is very large the change in ks,s-d becomes negligible due to the change in s and n, and therefore it can be treated that the value of k is constant.

Again the k can be shown to be a constant when span of earning life are assumed fixed regardless of schooling. Redefining n as a fixed span of earning life, we have:

[pic]……………………………..……….6

[pic]……………………..…….7

Since now we have the two expressions for Vs and Vs-d , the ks,s-d can be calculated by taking the ratio of:

[pic]………………………………….8

From the expression above it can be concluded that the earnings ratio k, of incomes differing by d years of schooling does not at all depend on the level of schooling s, nor on the length of earning life n, when that is finite or short.

If the conditions are changed so that s=d, that is s-d=0 we can define ks,0 as

[pic]………………………………………….9

Following the relation 8, the ks = ers…… ……………………………..10

Hence from equation 9 this relation can also be found [pic] ……………………………………………………….11

Taking the natural logarithm the formula 11 becomes

[pic]……………………………………………………12

The equation 12 gives the basic conclusion that the percentage increase in the earnings are strictly proportional to the absolute differences in the time spent at school, with the rate of return as the coefficient of proportionality, and therefore it can be precisely concluded that the logarithm earnings are strictly linear function of time spent in schooling.

2 Incorporating the Years in Schooling, Age and Sex in Earning Schooling Model

Investment to a person takes time that is a person to undergo training; in this case consideration of opportunity cost should be taken into account. Each additional period of schooling or job training results to postponement of time of the individual’s receipts of earning and therefore reduces the working life span (ibid).

The equation 12 can be expanded by incorporating post schooling investments in an econometric analysis of the distribution of the earnings so as to be able interpret age (experience) earning profile as consequences of investment behavior.

Experience profiles of log earnings are much more nearly parallel than the age profile, for this reason the earnings function in which the earnings are logarithmic, years of work experience should be entered additively and arithmetically. In this case the experience term is not linear but concave, and therefore the earning function is parabolic in the experience term (Mincer 1974). By considering the above argument the equation 12 becomes;

[pic]……………………………………13

Where t is years of experience and Ys is the earning capacity after completion of the schooling, by substituting equation 12 into 13 we obtained the following equation below,

[pic]………………………………14

For this case work experience is continuous and starts immediately after completion of schooling therefore t = (A-s-b). The factors which influences the experience term to be non linear include ‘the fitness of life, increasing the incidence of illness at older ages, and the secular progress of knowledge, which makes older education and skill vintages obsolescent, are convincing facts suggesting that as age advances, effects of depreciation eventually begin to surpass gross investment’[3].

Overview of determining the private economic return to education

To deal with different research questions, the following assumptions were put forward one is that all individuals started their schooling from pre-school level, no repeaters during training period. To complete primary school takes 9 years, 2 years for pre-school and 7 years for primary school. The second assumption is that all individual incurred zero cost of education regardless their location differences in the country, this is because data concerning the unity cost for education were not available. Also individuals of the same education level have the same marginal returns of their schooling regardless their differences in professional.

Return to education for different years were estimated by using OLS technique based on Mincer’s log earnings. Basing on the Schooling model the following regression equation was derived [pic]

Where, Y = an individual earning which is the dependent variable

[pic] = a constant

[pic] = highest level of schooling

[pic] = Marital status, dummy for marital status was constructed as 1 for marriage and 0 for non marriage

[pic] = Gender factor was captured and dummies were constructed as 1 for male while 0 for female

[pic] = Age of an individual

Age2 = Age *Age of and individual

[pic] = Urban or rural, urban was recorded 1 while rural 0

[pic] & [pic] are coefficients to be estimated

β is typically interpreted as the rate of return to an additional year of schooling with the assumption that the cost of education is zero. Education is the measure of highest level of education attended by individual.

CHAPTER FOUR: Summary Statistics

4.1 Data description for the year 1991/1992

The education statistics presented below based on our sample of 1991/91 as those used to analyze the private return to education basing on different levels of education as indicated in table 4.1

Table 4.1: Education statistics for the year 1991/1992

| Education Categories |1991/1992 Percentage s per Category |

|  |Men |Women |Total |

|No education |13.78 |35.26 |16.96 |

|Incomplete primary |21.61 |16.91 |20.91 |

|primary education |33.95 |25.87 |32.76 |

|Incomplete secondary |8.73 |6.50 |8.27 |

|secondary education |7.58 |5.64 |7.29 |

|Post secondary |14.35 |9.83 |13.81 |

|Diploma |  |  |  |

|University |  |  |  |

|Adult education |  |  |  |

|  |  |  |  |

|Number of observation | 3,903 | 684 | 4,587 |

Figure 4.1: Education Statistics for the year 1991/1992

In the table 4.1 and the figure 4.1 above, almost 17 per cent of Tanzanian had no education while 20 per cent had education level of at least secondary level. This implies that 80 per cent of population had education below secondary level including with no education. In general women had low level of education; over one third (35%) had no education as compared to only 13.78 per cent of men who had no education. In every level of education women had lower percentage compared to men except in zero education. Therefore men are better off because at least higher per cent of men had education up to secondary level and above which was about 21 per cent of men. In total large percent of population had their highest education level of primary education, this is due to education system in Tanzania where an individual to continue with further studies in secondary education he/she has to pass the primary examination. Incomplete primary education comprises almost 20 percent of the total sample while incomplete secondary education comprises 8 percent

4.2 Data description for the year 2000/2001

Table 4.2:- statistical data description 2000/2001

|Educational  Category |Men (%) |Women (%) |Total (%) |

|No education |12.28 |25.95 |18.04 |

|Incomplete primary |18.27 |15.48 |17.1 |

|primary education |44.63 |43.36 |44.09 |

|Incomplete secondary |5.62 |2.9 |4.46 |

|secondary education |8.87 |6.3 |7.79 |

|Post secondary |7.62 |5.29 |6.64 |

|Diploma |1.43 |0.56 |1.06 |

|University |1.28 |0.17 |0.81 |

|  |100 |100.01 |99.99 |

|Number of observation | 16,974 | 12,382 | 29,356 |

Figure 4.2: Summary Statistics for 2000/2001

In the 2000/2001 (table 4.2 and figure 4.2) the sample size used to analyze data included 29,356 individuals with 57.82 per cent men while women were 42.18 per cent. From that sample 18 per cent of population had zero education while 44 per cent had education level only up to primary education. Incomplete primary education comprises 17.1 per cent and incomplete secondary comprise 4.46 per cent. Individuals with education from secondary up to university reached 8.5 per cent, which means 91.5 per cent of population, had education level below secondary level in 2000/2001.

By considering men and women separately, from table 4.2 above there are 16,974 number of observation for men while there are 12,383 women. Out of 16,974 of men 12.28 per cent had zero education while 44.63 per cent had primary education level. Incomplete secondary education comprise 14.489 per cent while secondary is 8.87 per cent of the total number of observation for men. Above secondary education there was 10.33 per cent, there are great improvement of number of observation with individual with higher level of education, because from data of 1990/1991 no observation was found with individual of higher education.

On the other hand majority of women had very low level of education compared to men. Women with zero education were 25.28 per cent a difference of 13 per cent was observed between men and women. It means women are twice as much as men to have no education. This is a big challenge to the education for all policy because it seems that education for all begin with men and then women. But the Household budget survey 2000/01 (2002 pg. 36) indicates that girls have higher primary enrollment ratio than boys especially in rural areas were most of Tanzania live while higher enrollment for boys in urban area were just small percent about 35 per cent of population found. For normal situation it is difficult to understand the reason for higher number of women with zero education, may be the situation will be reversed after a number of years but apparently the results do not march with the HBS. 43.36 per cent of women had education up primary level here the difference is just 1.27 per cent. Incomplete secondary comprise only 2.9 per cent, a difference with men is 5.97 per cent is observed while in secondary level 6.30 per cent of women had a secondary education and the difference is 2.57 per cent. Higher education comprises 6.02 per cent of women here the difference is 4.31. In most cases men experience more per cent in different level of education unlike the women who’s their per cent outweigh men only in zero education.

By comparing women and men, men had higher level of education in all levels of education, for example 10.33 percent of men had at least a post secondary education while women were only 6.02 percent. Large percentage of women had no education which accounted to 25.95 percent. The percent of women without education decreased from 35.26 percent in 1991/92 to 25.95 percent by 2000/2001 at the same time the percentage of women with primary education increased in primary education from 25.87% in 1991/92 to 43.36% in 2000/2001. Also the percentage of men without education decreased from 13.78 percent in 1991/1992 to 12.28 per cent in 2000/2001

4.3 Data Description in 2006/2007

Table 4.3 Data Description in 2006/2007

|  |Men |Women |Total |

|No education |11.11 |29.52 |15.87 |

|Incomplete primary |13.13 |13.09 |13.12 |

|primary education |52.44 |40.69 |49.4 |

|Incomplete secondary |4.06 |2.78 |3.73 |

|secondary education |9.25 |5.64 |8.32 |

|Post secondary |5.88 |5.33 |5.74 |

|Diploma |1.26 |0.78 |1.14 |

|University |1.65 |0.74 |1.42 |

|Adult education |1.22 |1.41 |1.27 |

|  |100.00 |99.98 |100.01 |

|Number of observation | 7,303 | 2,547| 9,850 |

Figure 4.3: Data Description for 2006/2007

According to the table 4.3 and figure 4.3 the number of observation was 9850 with 7303 men and 2547 women which equals to 74.16 percent of men while women were 25.84 per cent. The average earning per month for men is much as women by more than two times. In general women had low level of education compared to men. About 29.52 of women had a zero level of education while men were only 11.11 per cent. Standard one to six comprise 13.09 per cent of women while men are 13.13 per cent. Primary level comprise a big per cent of total population because only few who pass the standard seven can continue with secondary education except for some private school which enroll even students with low marks. 52.44 per cent of men had a primary education while the percentage of women with primary education was 40.69, the difference of 11.75 was observed. Incomplete secondary comprise few individuals because this group may be of those who dropped in between due to various reasons including pregnancy for women or poverty which could a source of failure to meet the education costs. The incomplete secondary comprise 4.06 per cent of men and 2.78 per cent of women. For secondary education, about 9.25 per cent of men had education of secondary level while women were only 5.64 per cent. For post secondary men and women almost range on the same percentage which is 5.88 for men and 5.33 for women. Diploma is 1.26 per cent for men and 0.78 for women while university level men were 1.65 per cent and women were 0.74 per cent. Here the percentage of female students had increased followed the pre- entry program for women with lower qualification (Tanzania- poverty and Human development report 2009). Pre- entry program was introduced in 1999 with aim of increasing female student’s enrollment in higher learning institutions. Adult education is some sort of formal education where adult learn how to read and write but it is out of education system although ministry of education and vocational training facilitate the program.

Table 4.4:- Education categories for 1991/1992, 2000/2001 and 2006/2007 presented in a summarized table for comparison purpose

| |1991/1992 |2000/2001 |2006/2007 |

|  |

|Education levels |No,of Years |Male |Female |Total |

| |  |Return |Std err |Return |

| Education Levels |  |Return |Std err |Return |

| Education levels |  |Return |Std err |

  |Male |Female |Total |Male |Female |Total |Male |Female |Total | |incomplete primary |10 |12 |10 |6 |9 |7 |6 |4 |5 | |primary level |3 |10 |4 |12 |11 |12 |12 |11 |12 | |Incomplete secondary |10 |55 |20 |14 |24 |16 |11 |22 |13.1 | |secondary |30 |-96 |8 |20 |27 |24 |29 |18 |26.3 | |Advanced/post |7 |41 |8 |-11 |-42 |-18 |2 |18 |5.4 | |Diploma |  |  |  |17 |59 |28 |5 |17 |6.5 | |Tertiary |  |  |  |64 |59 |53 |118 |62 |108.2 | |

Figure 6.1 Trend of Returns to education from 1991/1992 to 2006/2007

[pic]

Basing on table 6.1 above, the return to incomplete primary education was decreasing for both male and female. In total the return had the following trend in 1991/1992 was 10 percent, in 2000/2001 was 7 percent and declined to 5 percent in 2006/2007. The return to women also shows the same trend in 1991/1992 an individual with incomplete primary education level appeared to be well rewarded for the one year course than complete primary or secondary education. But in 2000/2001 the situation changed and the return to incomplete primary education decreased from 12 percent to 9 per cent whereas by 2006/2007 the return had fallen to 5 per cent.

In 1991/1992 the return for men was higher in secondary education while for women was higher in incomplete secondary. Higher return were observed in 2000/2001 and 2006/2007 to higher education levels (tertiary) 64 per cent and 118 percent for men respectively at the same time as women had 59 by 62 per cent. Therefore the return to tertiary education is increasing while of secondary remain stable and return in primary education is characterized by concave structure. Here it is important to note that coding of education in the year 1991/1992 household budget survey data did not contain a separate category for diploma and tertiary education, therefore the analysis of trend for tertiary education starts from the year 2000/2001.

Following the decrease in return to incomplete primary education, it is possible that increase in supply of individuals with incomplete primary education level has reduced the private return to education as demand failed to respond in the same magnitude as supply. According to theory of demand and supply increase in student enrollment rate will cause excess supply of jobseekers that one will exert a downward pressure of the price of labor. The lower demand for low-skilled labor may be attributed by decrease in demand for goods and services produced by low-skilled workers (Bhula-or and Kripornsak 2008). Also technological progress leads to change in production process by acquiring high-skilled workers which are capable in using new and advance technology in production process (ibid)

The higher returns to post secondary level were observed by other studies like that of Schultz (2004) and Aromolaran (2004). The only difference with Schultz is that he obtained high return to post secondary education to women compared to men. Our result shows high return to men in higher education levels. This low return to women may be contributed by low supply of labor force in labor market. The result also was not consistence with Psacharopoulos (2002) who found that the return to schooling for a large number of countries were higher in primary, while it decreases as the number of year in schooling increases. In general, high return to tertiary is associated with low supply of population with high level of education and increase in demand for higher educated workers in labor market (Schultz 2004). Schultz argued that returns to higher schooling is likely to rise when the demand for better educated workers increase at the same that time the supply of better educated workers grows more slowly than the derived demand, possibly due to change in technical skills (ibid). A systematic change in the production process coupled with changes in technology leads to changes in the demand for certain type of labor (Psacharopoulos 2002), therefore highly educated individuals act as a complement to the higher level of technology.

Becker (1964) also noted that most highly educated workers almost tend to earn more than those of less education which means inequality in the distribution of earnings is positively related to inequality in education level. Although sometimes this may be invalid especially to individual with high experience they may earn more than higher educated worker. Also Söderbom et al, (2005) findings support Becker’s argument and our results that higher marginal return to education were found on post-secondary education compared to secondary level.

The challenge faced by society of having higher returns to higher levels of education is the increase in inequality between different levels of education. For rich families it will help to invest more to higher level of education while the problem will remain to the poor families who by majority will not be able to educate their families up higher levels. And Schulz (2004) argues that if private return to schooling increase at higher levels of education, poor families who are on average schooling their children up to primary school level will face low returns, while richer families who have ability to educate their children up to secondary level will face more return which will increase the gap between rich and poor families

By considering urban rural differences, the returns to education vary between rural and urban areas due to availability of labor market in urban areas as compared to rural. Also the imperfect mobility of labor between rural and urban may cause differences in return to education between the two regions. From the analysis individuals who are living in urban areas earn more income compared with those in rural areas. In 1991/1992 individuals who were living in urban area had a return of 72.13 per cent more than those living in rural areas. In 2000/2001 the gap between rural and urban narrowed down by 1.18 per cent because in 2000/2001 urban person earned 70.95 per cent more than that of rural areas. But in 2006/2006 data set did not contain the urban rural component, therefore the analysis containing urban rural were only for 1991/1992 and 2000/2001

CHAPTER SEVEN: Summary and Conclusion

This study was motivated by the education reform policy of 1995 and the increase in percentage of national budget on educational expenditure. The government has been putting more emphasis in primary education level as large percentage of the education sector budget is directed to finance primary education. However, it was inadequately known about the return to different levels of education, therefore this study provided estimates of the private returns to education for the year 1991/1992, 2000/2001 and 2006/2006. The return to education was estimated by using data from Tanzanian Household budget survey which may be the best sources of national statistical data. However the NBS collected data in order to measure the country’s poverty level but different indicators of poverty were also useful in analyzing the return to education like individual earnings in relation to education level although, the data concerning cost of education to each level of schooling were not collected. Therefore the effects of assuming zero cost of education may overstate the rate of return to a certain level of education hence the obtained estimates are termed as gross private returns. Also the individual earnings assumed to depend on his/her school participation and not on school choice or quality or other family background. The assumption helped to control for individual differences resulted from different family background as pointed out by Segundo (2003).

By using log earnings from Mincer (1974) model, the regression results confirm that all coefficients at all education level categories were positive and statistically significant at one percent in all three years. Hence, the results validate that there is a positive relationship between an individual’s level of schooling and his/her subsequent earnings (Mincer 1974). The results show that the returns to schooling tend to increase as average schooling level increases and the differences appear to have increased from 2000/2001 to 2006/2007. The returns to incomplete primary education and incomplete secondary education level were declining while the return to tertiary education was increasing. The reason behind for increasing return to tertiary education was very few people who had tertiary education therefore the results suggest that more attention should be given to higher education level because primary education has been universalized.

In order to study the female-male differences in return to education, the earning functions were estimated separately for male and female subsamples and the result showed that on average, women had lower returns to earnings compared to men and their return to education in almost every education category was lower than that of men except on 1991/1992 return for incomplete secondary education which was higher than that of men. The major factor which mentioned by more studies including Schultz (1988: 602) is that of low women participation in the labor market due multiple role of women including that of being much involved in unpaid family work. Another factor was; large percentage of women had low level of education including without education compared to men. By looking the trend of income inequality between men and women, there is a decline in income gap although the gap is still high because in 2006/2007 men earned on average 20 percent higher than women from 86.7 percent in 2000/2001. The trend was somehow convex since in 1991/1992 the gap was almost 60 percent, which was lower than that of 2006/2007 by 26.7 percent.

By considering urban-rural differences, the results confirm that there were higher returns to education in urban areas compared to rural. The dummy variable for urban was positive and statistically significant at one percent level of significance suggesting that earnings are significantly higher in urban areas compared to rural areas. In general individuals who were living in urban earned more than that of rural by 68 to 70 percent. By regressing men and women separately, the difference in return was higher for women than men. The difference in returns between urban and rural for women range from 73 to 74 percent while men range from 67 to 68 percent

By analyzing the trend of return to education, it was observed that trend of return to incomplete primary schooling and incomplete secondary education is declining indicating declining in demand for incomplete education categories. The decrease in return to incomplete primary education may be due to increase in supply of highly educated workers as employers increasing confidence on the categorization function of schools through examination results which lead to exclusion of individuals with less education (Solga 2002). The trend of return to tertiary education is increasing (consider figure 7.1 below) indicating the increase in difference in individual earnings based on education level.

Figure 7.1: Trend of Return to education from 1991/1992 to 2006/2007

[pic]

Finally, the trend of return to education suggests that the education expansion has resulted to decrease in return to lower level of schooling which resulted by increase in supply of the least educated individual in the economy while the demand shifted to highly educated individuals. And this may be true on the underlying assumption that the faster enrolments grow at one educational level relative to another will lower the rate of return on that investment (Psacharopoulos 1973). Hence, following the higher return to tertiary education, investment in tertiary education level should be an important feature of education policy because it is the level which yield more return in the labor market through higher earnings. Therefore government has to find a way of increasing its revenue so as to finance the expected increase in education expenditure as a result of increase in investment in tertiary education.

References

Aromolaran A.B. (2004), Wage Returns to Schooling in Nigeria. Africa Development Bank.

Basu, K. (1999) Child labor: Cause, Consequences, and Cure, with remarks on International Labor Standards. Journal of economic Literature, Vol. 37, No. 3 , pp 1083-1119

Bedi, A. S. and J. H.Y. Edwards., 2002. The Impact of School Quality on Earnings and Educational Returns - Evidence from a low- income country. Journal of Development Economics, Vol. 68 (2002), pp.157- 185.

Becker G. S. (1964). Human capital: A Theoretical and Empirical Analysis, with special reference to education

Bellay, C. (1999) Education for all. The state of the World Children: United Nations Children’s Fund (unicef): accessed on 1st October 2010

Bhula-or R. and P. Kripornsak (2008). Trends of Labour Demand for Higher- skilled workers and their wages in Thailand: Journal of International Business Research, Volume 7, special issue 1, 2008. Retrieved on 6th November 2010

Buchert L. (1997), Education Policy Formulation in Tanzania: Coordination between the Government and International aid Agencies

Card D. and A. Krueger (1996). Labour Markets Effects on School Quality: Theory and Evidence. In Burtless, Gary ed. Does Money matter? The Effect of School Resources on Students Achievement and Adult Success. Washington D.C Brookings Institution, pg. 97-140

Diaz O. JL (2008). The Economic Returns to Education in Mexico: a Comparison between Urban and Rural. Retrieved on 13rd October 2010

Fallon P. and D. Verry (1988) “The Economics of Labour Markets”

Gray H. A., (2003), Internal Rate of Return, Computer word Feb 17 2003, 37, 7, ABN/INFORM Global pg 32

Griliches Z. (1988): Technology, Education and Productivity. Early papers with Notes to Subsequent Literature

Gujarati D.N and D.C. Porter (2009): Basic Econometrics, Fifth Edition

Harmon, C., I. Walker and H. Oosterbeek (2003), The Returns to Education: Microeconomics. Journal of Economic surveys, Vol.17, No. 2

Heckman J.J, L.J Lochner and P. Todd (2003): Fifty Years of Mincer Regression. Retrieved on 07 Nov. 2010

http//tanzania.go.tz/education

Kahyarara G. and F. Teal (2006):- General or Vocational Education? Evidences from the Returns to Education in Tanzanian Manufacturing Firms.

26 October 2010



(

Retrieved on 15th October 2010

Retrieved on 9 November 2010

Retrieved on 27th October 2010

Maneschiöld P. and B. Haraldson (2007). Religious Norms and Labour supply of marriage women in Sweden: Finnish Economic Papers. 20, 1, ABN/INFORM Global

Mincer J. (2003) Technology and Labor Market: Review of Economics of Households; Dec 2003; 1,4 ABI/INFORM Global pg 249-272

Mincer J. (1974) Schooling, Experience and Earnings: National Bureau of Economic Research, New York pg 8

Ministry of Education and Vocational Training (2009):- Basic Education Statistics in Tanzania (BEST) 2005-2009, National Data

Ministry of Education and Vocational Training (2010):- Basic Education Statistics in Tanzania (BEST) 2006-2010, National Data

Mingat A., R. Rakotomalala and J.P. Tan (2002) Financing Education for All by 2015: Simulations for 33 African Countries. Retrieved on 16 October 2010

Murphy K.M and F. Wilch (1990). Empirical Age -Earning profile. University of Chicago

National Bureau of Statistics Tanzania (2002), Household Budget Survey 2000/2001

Psacharopoulos, G., and K. Hinchliffe (1973), Returns to Education. An International Comparison, Higher education Research Unit. London school of Economics and Political Science. Elsevier Scientific Publishing company Amsterdam

Psacharopoulos, G (1994), Returns to Investment in Education. A. global update. World development, pp. 1325-1343. The world bank Washngton Dc

Psacharopoulos, G. and H. A. Patrinos (2002). Returns to Investment in Education. Policy Research Working paper No. 2881.

Research and Analysis working Group,(2005). Poverty and Human Development report 2005

Schultz, T.P., (1988). Education Investment and Returns. In: H. Chenery and T.N. Srinivasan, eds. 1988. Handbook of Development Economics Volume 1. Amsterdam: Elsevier Science Publishers B.V: Chapter 13.

Schultz, T. P. (2004). Evidence on Returns on Schooling in Africa from Household surveys: Monitoring and Restructuring the Market for Education. Journal of Africa Economies, Volume 13, AERC supplement 2, pp. ii95-ii148

Seebens H.and P. Wobst (2005), The Impact of Increased School Enrollment on Economic Growth in Tanzania. African Development Bank



Segundo, M.J.S and A. Valiente (2003), Family Background and the returns to Schooling in Spain. Education Economics, Vo. 11, No. 1, 2003

Smith S. (1994, 2003) Labour Economics. 2nd Edition

Söderbom M., F. Teal, A. Wambugu and G. Kahyarara (2005):- The Dynamics of Returns to Education in Kenyan and Tanzanian Manufacturing*

Website:

Date 31 October 2010

Solga H. (2002). ‘Stigmatization by Negative Selection’. Explaining Less Educated Peoplel’s Decreasing Employment Opportunities. Vol. 18 No. 2, 159-178. Retrieved on 1st October 2010

The International Bank for Reconstruction and Development (1993): The East Asian Miracle. Economic Growth and Public Policy. World Bank, Oxford University Press

Thornton R.J., J.D. Rogers and M.L. Brookshire (1997). On the Interpretation of Age- Earnings Profiles: Journal of Labor Research Volume XVIII, Number 2

United Republic of Tanzania (2009): Poverty and Human Development Report (2009): Research and analysis working group, MKUKUTA Monitoring System. Ministry of Finance and Economic Affairs

United Republic of Tanzania (2000-2008): Millennium Development Goals Report, Mid- Way Evaluation

Wilmoth J. and G. Koso (2002) Does Marital History Matter? Marital Status and Wealth Outcomes Among Preretirement Adults Journal of Marriage and Family 64 (February 2002): 254–268 Purdue University

Appendices

Regression for year 1990/1991

[pic]

[pic]

[pic]

2000/2001

[pic]

[pic]

[pic]

1

2

2006/2007

[pic]

[pic]

[pic]

-----------------------

[1] Form one is the first year of secondary education, normally students starts secondary education after completing primary education. For public schools only who pass primary school leaving examination extend their schooling to secondary education while for private schools they set their own criteria for selection to form one.

[2] Children generate income by being involved in family work and even in child labor for poor family , that is child labor act as buffer stocks against income shocks (Basu, 1999)

[3] Mincer (1974, pg 20-21) Theoretical analysis; Individual Acquisition of earning Power

[4] as discussed in chapter four, women had lower level of schooling compared to men

-----------------------

RETURNS TO EDUCATION IN TANZANIA

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download